Computer-generated models of blood coagulation factor Xa, factor IXa, and thrombin based upon structural homology with other serine proteases

Surface plasmon resonance studies of the interaction between factor VII and tissue factor. Demonstration of defective tissue factor binding in a variant FVII molecule (FVII-R79Q)

The blood coagulation cascade is initiated when vessel injury allows factor VII (FVII) to form a complex with tissue factor (TF). Complete deficiency of FVII causes a lethal bleeding diathesis, but individuals with moderately reduced FVII levels are often asymptomatic. Some of these individuals have circulating partially functional FVII, as a result of point missense mutations in critical parts of the molecule. One such mutation has been reported at position 79 in the first epidermal growth factor-like (EGF) domain of FVII, where an arginine residue has been replaced by glutamine. There is controversy as to whether or not this mutation reduces the affinity of the FVII/TF interaction compared to wild-type FVII. To address this problem, we have expressed recombinant FVII-R79Q and subjected it to detailed biochemical analysis. One-stage FVII:C assays show the variant FVII to have reduced activity with respect to the wild type. Rates of autoactivation and activation by FXa to the two-chain molecule were identical for wild-type and variant FVII. The Vmax for FX activation was lower for the mutant as measured using an amidolytic assay for FX activity. In contrast, the Km for FX was lower for the variant than the wild-type molecule. Peptidyl substrate hydrolysis was virtually identical for both variant and normal FVIIa in the presence and absence of TF. The variant has reduced affinity for TF as measured by surface plasmon resonance. FVII-R79Q has an association rate constant (kassoc) one-fifth of that of normal FVII, but a similar kdiss, resulting in a decrease in the affinity of the enzyme for its cofactor.(ABSTRACT TRUNCATED AT 250 WORDS)

A structural model for the prostate disease marker, human prostate-specific antigen

Prostate-specific antigen (PSA) provides an excellent serum marker for prostate cancer, the most frequent form of cancer in American males. PSA is a 237-residue protease based on sequence homology to kallikrein-like enzymes. To predict the 3-dimensional structure of PSA, homology modeling studies were performed based on sequence and structural alignments with tonin, pancreatic kallikrein, chymotrypsin, and trypsin. The structurally conserved regions of the 4 reference X-ray proteins provided the core structure of PSA, whereas the loop structures were modeled on the loops of tonin and kallikrein. The unique "kallikrein loop" insert, between Ser 95b and Pro 95k of kallikrein, was constructed using molecular mechanics, dynamics, and electrostatics calculations. In the resulting PSA structure, the catalytic triad, involving residues His 57, Asp 102, and Ser 195, and hydrophobic and electrostatic interactions typical of serine proteases were extremely well conserved. Similarly, the 5-disulfide bonds of kallikrein were also conserved in PSA. These results, together with the fact that no major steric clashes arose during the modeling process, provide strong evidence for the validity of the PSA model. Calculation of the electrostatic potential contours of kallikrein and PSA was carried out using the finite difference Poisson-Boltzmann method. The calculations revealed matching areas of negative potential near the catalytic triad, but differences in the positive potential surrounding the active site. The PSA glycosylation site, Asn 61, is fully accessible to the solvent and is enclosed in a positive region of the isopotential map. The bottom of the substrate specificity pocket, residue S1, is a serine (Ser 189) as in chymotrypsin, rather than aspartate (Asp 189) as in tonin, kallikrein, and trypsin. This fact, plus other features of the S1 binding-pocket region, suggest that PSA would prefer substrates with hydrophobic residues at the P1 position. The location of a potential zinc ion binding site involving the side chain of histidines 91, 101, and 233 is also suggested. This PSA model should facilitate the understanding and prediction of structural and functional properties of this important cancer marker.

Biochemistry of factor X

Factor X circulates as a serine protease which is converted to the active form at the point of convergence of the intrinsic and extrinsic coagulation pathways. Subsequently, the enzymatic species, factor Xa, is involved in macromolecular complex formation with its cofactor factor Va, a phospholipid surface and calcium to convert prothrombin into thrombin. The gene encoding factor X shares a number of structural and organisational features in common with the other vitamin K-dependent coagulation proteins, suggesting that they have evolved from a common ancestral gene. Each of the exons encoding these proteins can be considered as a module coding for a homologous domain in each protein. These structural domains in factor X are responsible for specific functional properties including gamma-carboxylase recognition, calcium binding, phospholipid surface interaction, as well as cofactor and substrate binding. Studies of recombinant proteins and proteolytic fragments continue to provide significant insight into structure-function relationships of the protein modules within factor X.

Molecular defects in CRM+ factor VII deficiencies: modelling of missense mutations in the catalytic domain of FVII

The molecular defects causing CRM+ factor VII deficiency were investigated in seven unrelated subjects and several members of their families. Four missense mutations located in the catalytic domain of factor VII were found. The previously reported 304Arg-->Gln substitution was present in the homozygous and heterozygous forms, with different polymorphic haplotypes, thus demonstrating that it is recurrent and frequent in the Italian population. The 310Cys-->Phe substitution was found in the homozygous form and in the compound heterozygous condition with the nonsense mutation 356Trp-->stop. Two missense mutations, 298Met-->Ile and 342Gly-->Arg, were found in the homozygous and in the heterozygous condition respectively. Molecular heterogeneity was further increased by finding of the 353Arg-->Gln polymorphism in the doubly heterozygous condition with the 304 and 342 mutations. Plausible explanations for loss of FVII function were found by inspecting a model of the serine protease domain of factor VIIa. Inefficient activation of the catalytic site is predicted for 298Met-->Ile. 342Gly-->Arg would directly distort the geometry of the 'oxyanion hole' preventing formation of a substrate enzyme intermediate. 310Cys-->Phe is predicted to have an adverse effect on tissue factor interaction. These mutations point to important regions of the factor VII molecule.

Analysis of the partial nucleotide sequences and deduced primary structures of the protease domains of mammalian blood coagulation factors VII and X

In order to obtain sequence data for the blood coagulation factor VII and factor X in several mammalian species, we amplified and sequenced the DNA segments of exon VIII from each gene by means of the polymerase chain reaction (PCR) method. The DNA segments from the following species were successfully amplified: factor VII from the rhesus monkey and dog, and factor X from the rhesus monkey, Syrian hamster and rat. In each factor, the nucleotide sequences and predicted primary structures of the protease domain showed a high degree of homology among species; amino acid identities of approximately 68%-92% and 80%-98% were demonstrated among species in factor VII and factor X, respectively. The locations of the active site residues and five Cys residues were evolutionarily conserved in both factors. Interestingly, the amino acids involved in the human genetic variants, both factor VII 304-Arg and factor X 326-Arg, were always conserved across species. The data presented here will be helpful for investigating human genetic variants of factor VII or X, and will provide considerable information for constructing in vitro site-specific mutants of these factors.

Knowledge-based protein modeling

Knowledge, both from the three-dimensional structures of homologous proteins and from the general analysis of protein structure, is of value in modeling a protein of known sequence but unknown structure. While many models are still constructed at least in part by manual methods on graphics devices, automated procedures have come into greater use. These procedures include those that assemble fragments of structure from other known structures and those that derive coordinates for the model from the satisfaction of restraints placed on atomic positions.

Specific molecular interaction sites on factor VII involved in factor X activation

Factor VII, a serine-protease zymogen, and tissue factor, the cellular receptor/coenzyme, are the protein components of the macromolecular complex which initiates the extrinsic pathway of the coagulation cascade. Previous studies were directed to the identification of functional sites on factor VII which mediate factor X activation, employing a series of potentially inhibitory synthetic peptides representing the primary structure of factor VII and antibodies to selected peptides. The involvement of at least four high-affinity interactive regions [factor VII (44-50), (196-229), (285-305) and (376-396) peptides] on the surface of factor VII was clearly demonstrated. The minimal sequences for the expression of inhibitory activity of these four molecular recognition domains on factor VII were identified using short and overlapping peptides. The short factor VII-(206-218)-peptide (most inhibitory peptide in the sequence 196-229 on factor VII) inhibited the binding of factor VII to the tissue-factor-expressing J82 cell line. Furthermore, radiolabeled [Tyr201] factor VII-(199-221)-peptide, with a tyrosine substituted for the normal tryptophan residue, was specifically bound to J82 cells, and also the binding of the radiolabeled peptide to this cell line was specifically inhibited by a monoclonal antibody to tissue factor, confirming that the interaction site for tissue factor on factor VII is present within the peptide sequence 196-229. Kinetic analyses suggested that the regions represented by factor VII-(285-305)- and factor VII-(376-396)-peptides are involved in factor X recognition and the chemical cross-linking of the radiolabeled peptides resulted in specific binding to factor X, confirming that these two regions on factor VII represent the substrate-recognition site. Furthermore, these radiolabeled peptides specifically interact with the heavy chain of factor X, suggesting that the complementary binding region for the substrate-recognition site on factor VII are present on the heavy chain of factor X.

Symptomatic type II protein C deficiency caused by a missense mutation (Gly 381-->Ser) in the substrate-binding pocket

A patient with recurrent deep vein thrombosis and heterozygous type II deficiency, characterized by reduced protein C activity in both amidolytic and clotting functional assays, was investigated by direct sequencing of PCR fragments derived from the coding portion of the protein C gene. AG (8856) to A transition was noted in the patient which was not present in healthy controls. This mutation is predicted to cause the substitution of Ser for Gly 381, an evolutionari'y conserved residue in the substrate binding pocket of serine-proteases (Gly 216, chymotrypsin numbering). A computer model of the structure of the serine-protease domain indicates that the properties of the altered protein C molecule can be explained on the basis of steric hindrance between the substituted serine and the substrate arginine side chains.

A molecular model of the serine protease domain of activated protein C: application to the study of missense mutations causing protein C deficiency

A molecular model of the serine protease domain of protein C was constructed by standard comparative methods. Individual missense mutations were inserted into the model and plausible explanations for their interference with protein C structure/function were derived through consideration of location, steric effects and protein stability. A hydrophilic cluster of many Arg and Lys residues, found adjacent to the active site cleft, is proposed to be involved in thrombomodulin and/or protein S interactions. Analysis of comparative binding studies also suggested the presence of an extended substrate binding pocket in the model.

The dysfunction of coagulation factor VIIPadua results from substitution of arginine-304 by glutamine

This study addresses whether a mutation in the factor VIIPadua gene could explain the reduced activity of the inherited variant protein. All nine exons of the normal and Padua factor VII gene were amplified using the polymerase chain reaction, cloned into pUC19 and sequenced. A point mutation (G to A at nucleotide position 10828) was found which results in the substitution of a glutamine (CAG) for arginine (CGG) at amino acid position 304. This substitution creates a PvuII restriction site useful in screening for the defect and in demonstrating homozygosity. This substitution involves an arginine residue in the catalytic domain within a Leu*****Pro******Cys motif which occurs in conserved region 5 in up to 16 coagulation and other serine proteinases. On the basis of conformational homology among serine proteinases, it is suggested that the observed amino acid substitution in factor VIIPadua could cause structural changes affecting its activation and/or catalytic activity.

FVIIa derivatives obtained by autolytic and controlled cathepsin G mediated cleavage

The heavy chain of coagulation factor VII contains a serine esterase entity. A partial cleavage in the heavy chain occurs during purification and activation of the single-chain zymogen, presumably as a result of autolysis. Neutrophil cathepsin G initially generates a Gla-domainless FVIIa without coagulant activity. However, on extended exposure cleavage also occurs in the heavy chain, resulting in a complete loss of enzyme activity. Four cleavage sites on the heavy chain, two susceptible to trypsin-like autolysis and two susceptible to chymotrypsin-like cathepsin G-mediated catalysis have been identified. The hydrolysis of peptide bonds in the heavy chain might contribute to regulation of the coagulation process in vivo.

Antibodies to recombinant fragment 212-276 of protein C specifically recognize the intact human molecule

The peptide fragment Pro212-Ile276 of human protein C was produced as a part of a fusion protein in Escherichia coli. The identity of the peptide was confirmed by immunoblotting experiments using specific antibodies to intact protein C. The peptide Pro212-Ile276 was isolated from the fusion protein after mild hydrolysis with formic acid by gel filtration and reverse-phase HPLC. This peptide fragment was used to produce antibodies specific for the heavy chain of protein C which recognized native protein C present in blood plasma. Antibodies to intact protein C reacted also with the Pro212-Ile276 peptide fragment, indicating that this region is immunogenic in intact protein C and may represent a native epitope.

Single-strand conformation polymorphism (SSCP) analysis of the molecular pathology of hemophilia B

In the present study, we report the application of polymerase chain reaction-single-strand conformation polymorphism (PCR-SSCP) analysis to the screening of seven functionally important factor IX gene (FIX) regions (total length 2.66 kb) in 9 unrelated haemophilia B patients of Portuguese or African origin. In eight of the patients an altered migration pattern of single-stranded DNA was observed. Direct sequencing of the relevant DNA fragments unveiled the following sequence alterations: two novel mutations, namely FIXBarcelos Thr-380-Pro and FIXLousada 9bp insertion at position 31,309 or 31,318; five mutations previously reported in other ethnic groups (FIXPorto Arg-145-His, FIXLuanda Gly-207-Arg, FIXPenafiel Arg-248-Gln, FIXSesimbra Arg-333-Gln, FIXCascais Arg-333-Stop); and a normal variant, G-->T transvertion at position 6,596 in intron 2. We propose hypothetical models for the generation of the 9 bp duplication (FIXLousada). We have performed molecular modeling studies in order to predict the structure of the variant FIX molecules.

Buried water in homologous serine proteases

Buried water molecules in the structurally homologous family of eukaryotic serine proteases were examined to determine whether buried waters and their protein environments are conserved in these proteins. We found 16 equivalent water sites conserved in trypsin/ogen, chymotrypsin/ogen, elastase, kallikrein, thrombin, rat tonin and rat mast cell protease, and 5 additional water sites in enzymes which share the primary specificity of trypsin. Based on an alignment of 30 serine protease sequences, it appears that the protein environments of these 21 conserved buried waters are highly conserved. The protein environments of buried waters are comprised primarily of atoms from highly conserved residues or main chain atoms from nonconserved residues. In one instance, the protein environment of a water is conserved even in the presence of an unlikely Pro/Ala substitution. We also note 3 instances in which a histidine side chain substitutes for water, suggesting that the structural role of water at these sites is satisfied by the presence of an alternative hydrogen bonding partner. Buried waters appear to be integral structural components of these proteins and should be incorporated into protein structures predicted on the basis of sequence homology to this family, including the catalytic domains of coagulation proteases.

Primary structure of the virus activating protease from chick embryo. Its identity with the blood clotting factor Xa

Host cell proteases activating para- and orthomyxovirus fusion glycoprotein precursors play a crucial role in determining the viral tropism in infected organisms. We previously isolated such an endoprotease from the allantoic fluid of chick embryo and showed its close similarity to the activated form of blood clotting factor X (FXa) by partial amino acid sequencing. In this report, we have cloned and sequenced a cDNA of the protease, and show that it is encoded in a single gene as a preproform with all the functional and structural domains known to be characteristic of bovine or human FX, establishing the identity between the protease and FXa.

Model structure for the human blood coagulation agent beta-factor XIIa

An improvement to the human blood coagulation agent beta-factor XIIa three-dimensional model is proposed. The sequence alignment as well as the modeling procedures are presented and the minimized energy of the new model is reported before and after solvation of the active center.

Prothrombin mRNA is expressed by cells of the nervous system

Thrombin, a serine protease of the blood coagulation system, has additional effects on cells in vitro. It is mitogenic for fibroblasts and astrocytes and contributes to the regulation of neurite outgrowth and astrocyte stellation. Until now the expression of thrombin or its precursor prothrombin in tissues other than liver has not been demonstrated conclusively because of difficulty in avoiding serum contamination. Using sensitive mRNA detection methods, we show here that prothrombin is expressed not only in the liver, but also in the brain throughout development. Polymerase chain reaction, Northern, and in situ hybridization studies demonstrate the presence of prothrombin transcripts in the olfactory bulb, the cortex, the cerebellum, and other regions of the rat and human nervous system, as well as in neural cell lines. These results support an involvement of (pro)thrombin in the regulation of cellular events in the nervous system.

Primary structure of rat hepatocyte growth factor and induction of its mRNA during liver regeneration following hepatic injury

Overlapping cDNA clones for rat hepatocyte growth factor (rHGF) were isolated by cross-hybridization with the cloned cDNA for human hepatocyte growth factor (hHGF) and the nucleotide sequence of the cDNA was determined. The entire primary structure of rHGF was deduced from the sequence. Comparison of the amino acid sequences between rat and human HGFs revealed that the two sequences are highly conserved throughout the protein structures, suggesting that rat and human HGFs may be functionally similar. Responses of the rHGF mRNA during liver regeneration in rats were examined by Northern blot hybridization analysis with the aid of the cDNA probe for rHGF. The mRNA levels increased in the liver and spleen but not in the kidney after administration of carbon tetrachloride. At the maximum level of induction, the rHGF mRNA increased in the liver about 4.5-fold over its normal level. The mRNA levels also increased in the liver and spleen after administration of D-galactosamine. On the other hand, no obvious increase of the mRNA was observed in the liver and spleen after partial hepatectomy. These observations suggest that HGF may function as a regulator of liver regeneration following hepatic injury caused by hepatotoxins.

A critical evaluation of the predicted and X-ray structures of alpha-lactalbumin

The rapidly increasing availability of protein amino-acid sequences, many of which have been determined from the corresponding gene sequences, has intensified interest in the prediction of related protein structures when the three-dimensional structure of another member of the family is known. The study of bovine alpha-Lactalbumin provides a classic example in which the three-dimensional structure was predicted, first by Browne et al. (1969) and later by Warme et al. (1974), from the three-dimensional structure of hen-egg-white lysozyme (Blake et al., 1965), taking into account the striking relationship between the amino acid sequences of the two proteins. A comprehensive comparison of these models with the structure of baboon alpha-Lactalbumin derived from X-ray crystallography (Acharya et al., 1989) is presented. The models mostly compare well with the experimentally determined structure except in the flexible C-terminal region of the molecule (rms deviation on C alpha s of residues 1-95, 1.1 A).

Protein modeling of human prorenin using the molecular dynamics method

To study the activation-inactivation mechanism of the renin zymogen, prorenin, a tertiary structural model of human prorenin was constructed using computer graphics and molecular dynamics calculations, based on the pepsinogen structure. This prorenin model shows that the folded prosegment polypeptide can fit into the substrate binding cleft of the renin moiety. The three positively charged residues, Arg10, Arg15, and Arg20, in the prosegment make salt bridges with Asp225, Glu331, and Asp60, respectively, in renin. Arg43, which is in the processing site, forms salt bridges with the catalytic residues of Asp81 and Asp269. These ionic interactions between the prosegment and the renin may contribute to keeping the prorenin structure as an inactive form.

Affinity labeling of lysine-149 in the anion-binding exosite of human alpha-thrombin with an N alpha-(dinitrofluorobenzyl)hirudin C-terminal peptide

In order to define structural regions in thrombin that interact with hirudin, the N alpha-dinitrofluorobenzyl analogue of an undecapeptide was synthesized corresponding to residues 54-64 of hirudin [GDFEEIPEEY(O35SO3)L (DNFB-[35S]Hir54-64)]. DNFB-[35S]Hir54-64 was reacted at a 10-fold molar excess with human alpha-thrombin in phosphate-buffered saline at pH 7.4 and 23 degrees C for 18 h. Autoradiographs of the product in reducing SDS-polyacrylamide gels revealed a single 35S-labeled band of Mr approximately 32,500. The labeled product was coincident with a band on Coomassie Blue stained gels migrating slightly above an unlabeled thrombin band at Mr approximately 31,000. Incorporation of the 35S affinity reagent peptide was found markedly reduced when reaction with thrombin was performed in the presence of 5- and 20-fold molar excesses of unlabeled hirudin peptide, showing that a specific site was involved in complex formation. The human alpha-thrombin-DNFB-Hir54-64 complex was reduced, S-carboxymethylated, and treated with pepsin. Peptic fragments were separated by reverse-phase HPLC revealing two major peaks containing absorbance at 310 nm. Automated Edman degradation of the peptide fragments allowed identification of Lys-149 of human thrombin as the major site of DNFB-Hir54-64 derivatization. These data suggest that the anionic C-terminal tail of hirudin interacts with an anion-binding exosite in human thrombin removed 18-20 A from the catalytic apparatus.

Binding of the bovine basic pancreatic trypsin inhibitor (Kunitz inhibitor) to human and bovine factor Xa. A thermodynamic study

The effect of pH and temperature on the apparent association equilibrium constant (Ka) for the binding of the bovine basic pancreatic trypsin inhibitor (BPTI, Kunitz inhibitor) to human and bovine factor Xa (Stuart-Prower factor; EC 3.4.21.6) has been investigated. Under all the experimental conditions, values of Ka for BPTI binding to human and bovine factor Xa are identical. On lowering the pH from 9.5 to 4.5, values of Ka (at 21.0 degrees C) for BPTI binding to human and bovine factor Xa decrease, thus reflecting the acidic pK shift of the His57 catalytic residue from 7.1, in the free enzyme, to 5.2, in the proteinase-inhibitor complex. At pH 8.0, values of the apparent thermodynamic parameters for BPTI binding to human and bovine factor Xa are: Ka = 2.1 x 10(5)M-1 (at 21.0 degrees C), delta G degree = -29.7 kJ/mol (at 21.0 degrees C), delta S degree = +161 entropy units (at 21.0 degrees C), and delta H degree = +17.6 kJ/mol (temperature-independent over the explored range, from 5.0 degrees C to 45.0 degrees C). Thermodynamics of BPTI binding to human and bovine factor Xa have been analysed in parallel with those of related serine (pro)enzyme/Kazal- and /Kunitz-type inhibitor systems. Considering the known molecular models, the observed binding behaviour of BPTI to human and bovine factor Xa was related to the inferred stereochemistry of the proteinase/inhibitor contact region.

Quenching of the amidolytic activity of one-chain tissue-type plasminogen activator by mutation of lysine-416

In contrast to most other serine proteases, tissue-type plasminogen activator (t-PA) possesses enzymatic activity as the one-chain zymogen form. The hypothesis that lysine residues 277 or 416 may be involved in stabilization of an active conformation of one-chain t-PA via salt-bridge formation with aspartic acid residue 477 was tested by site-directed mutagenesis. Four recombinant t-PA mutants were constructed. The amidolytic activities of these analogues were compared to that of authentic t-PA. Substitution of arginine-275 provided an analogue [( R275G]t-PA) resistant to plasmin cleavage. The amidolytic activity of [R275G]t-PA was comparable to that of authentic one-chain t-PA, and so was the activity of [R275L,K277L]t-PA, in which additional substitution of lysine residue 277 was carried out. This suggested that its presence was nonessential for obtaining one-chain t-PA activity. In contrast, substitution of lysine residue 416 to obtain [K416S]t-PA and [K416S,H417T]t-PA resulted in substantial quenching of amidolytic one-chain activity. As expected, the amidolytic activities of the two-chain forms were less affected by the substitution. Involvement of lysine residue 416 in one-chain t-PA activity was also indicated by decreased activities of [K416S]t-PA and [K416S,H417T]t-PA with plasminogen as the substrate. The one-chain activity of the lysine residue 416 substitution analogues was partially restored in the presence of fibrin. This could indicate that strong ligands such as fibrin might provide an alternative stabilization of the active conformation of one-chain t-PA.

Dominant and recessive alleles of the Drosophila easter gene are point mutations at conserved sites in the serine protease catalytic domain

The easter gene, required for the development of all lateral and ventral pattern elements in the Drosophila embryo, appears to encode an extracellular serine protease. Dominant easter alleles increase the number of cells that give rise to lateral and ventral structures. We have found that nine dominant and four recessive mutations are caused by single amino acid substitutions at conserved sites in the putative serine protease catalytic domain. The activity of dominant products was assayed by injecting in vitro synthesized transcripts from the dominant alleles into young embryos. The results suggest that the dominant easter products cleave the normal substrate, but fail to respond to a spatially asymmetric regulator.

The cDNA structure of rat plasma kallikrein

From a liver cDNA library we have isolated and characterized the cDNA encoding rat plasma kallikrein. The cDNA structure contains 2,456 nucleotides with a 2,082-nucleotide-long open reading frame. Protein sequence data suggest that the signal peptide is 19 amino acids long. This results in a mature plasma prekallikrein containing 619 amino acids. Determination of tissue distributions using Northern blot analysis (3.0-kb transcript) and the polymerase chain-reaction methodology on RNA preparations demonstrated that in the rat the liver is the main source of this enzyme. Southern blots suggested the presence of a single gene coding for rat plasma kallikrein. Finally, although Southern blots revealed a homologous gene in mouse, the mRNA corresponding to the mouse hepatic proteinase is barely detectable on Northern blots, suggesting inefficient transcription or high turnover of the mRNA in this species.

Conformational lability of vitronectin: induction of an antigenic change by alpha-thrombin-serpin complexes and by proteolytically modified thrombin

We previously showed that the alpha-thrombin-antithrombin III complex causes antigenic change in vitronectin as monitored by the monoclonal anti-vitronectin antibody 8E6 (Tomasini & Mosher, 1988). We have extended these studies to other protease-serpin complexes and to gamma-thrombin, a proteolytic derivative of alpha-thrombin. In the presence of heparin, recognition of vitronectin by 8E6 was increased 64- or 52-fold by interaction with the complex of alpha-thrombin and heparin cofactor II or the Pittsburgh mutant (Met358----Arg) of alpha 1-protease inhibitor, respectively. This was comparable to the value obtained with the alpha-thrombin-antithrombin III complex. Factor Xa-serpin complexes were approximately 4-fold less effective than the corresponding thrombin complexes. alpha-Thrombin-serpin complexes but not Xa-serpin complexes formed disulfide-bonded complexes with vitronectin. Antigenic changes and disulfide-bonded complexes were not detected when trypsin- or chymotrypsin-serpin complexes were incubated with vitronectin. gamma-Thrombin caused 7- and 34-fold increases in recognition of vitronectin by MaVN 8E6 in the absence and presence of heparin, respectively. In contrast, alpha-thrombin by itself had no effect. The antigenic change induced by gamma-thrombin was maximal when gamma-thrombin and vitronectin were equimolar, was not dependent on cleavage of vitronectin, and was abolished by inhibition of gamma-thrombin with Phe-Pro-Arg-chloromethyl ketone but not with diisopropyl fluorophosphate. These data indicate that alpha-thrombin is the component in alpha-thrombin-serpin complexes that induces the antigenic change in vitronectin, probably via a region that is preferentially exposed in gamma-thrombin.